Washington - With data from the Puerto Rico Demographic Registry -that the island´s government kept in secret-, two Puerto Rican researchers at Harvard University have estimated that in the 200-day period after Hurricane Maria, there were 3,433 more people than expected.
The analysis covers from September 20, 2017 - the date in which the hurricane hit the island - until April 15, 2018, when, according to data from the Demographic Registry, that mortality rate had normalized.
The analysis of Puerto Rican researchers Rafael Irizarry and Rolando Acosta, from the Harvard Department of Biostatistics, is consistent with the estimate of the George Washington University Milken Institute of Public Health (GWU), which concluded that deaths related to the emergency caused by Hurricane Maria were about 2,975, a total that the government of Puerto Rico accepted.
But, at the same time, it offers a look at the causes of 2,392 of the deaths that occurred between September 20, 2017 and December 31, 2017.
Irizarry and Acosta study argues that excess deaths may have been between 3,189 and 3,676, to reach a confidence interval of 3,433.
"If you stop the analysis on February 28, as GWU did, the estimate is almost the same," said Irizarry - professor at Harvard – no the telephone. Acosta is in the second year of the Harvard Biostatistics PhD program.
Without using the Demographic Registry data, researchers led by the Harvard School of Public Health released a survey in May, which estimated that there were 4,645 deaths caused by María, with an average between 793 and 8,498.
It was not until June 1 that the Demographic Registry disclosed the deaths that occurred between September 2017 and February 2018, after lawsuits filed by CNN and the Center for Investigative Journalism.
In fact, the GWU Milken Institute of Public Health indicated that its researchers were not given the Demographic Registry data until mid-May.
The new analysis of the Puerto Rican researchers at Harvard includes the data on the daily mortality rate offered by the Demographic Registry, changes in population and seasonal mortality trends.
The study - which has not been academically published yet - includes the statistical code they used, since they are willing to revise their estimates if, for example, they are presented with different data on the reduction of the number of inhabitants.
Irizarry said that they base the population estimate on data from the Puerto Rico Institute of Statistics and Teralytics company. In this sense, they agree with GWU's analysis that the island´s population drop to about 3 million, a loss of almost 300,000 people.
Professor Irizarry noted that they examined 30 causes of death, and that those that showed a significant increase tend to confirm the consequences of the prolonged lack of electricity and health services.
Out of the 2,392 excess deaths that occurred between September 20 and December 31, 2017, 703 were due to circulatory system problems, 330 to endocrine diseases, 289 to problems with the nervous system, 269 to respiratory complications and 161 to infections with bacteria.
"Only a few rose, and there are many deaths associated with problems in the healthcare system, such as lack of electricity. That includes deaths from endocrine diseases, bacterial infections - which tend to be a reflection of arriving late to the hospital - heart attacks that are consistent with an increase in stress. Other causes, such as viruses or murder, did not go up," said Irizarry, stressing that all of this" is evidence that after the hurricane, people did not have access to the healthcare services they had before the hurricane."
1,417 deaths due to Georges
In their analysis, the Puerto Rican researchers also calculated that total excess deaths in Puerto Rico after facing the fury of Hurricane Georges, in 1998, was 1,427.
Officially, the government of Puerto Rico maintains that there were no "direct" deaths related to that hurricane, which hit the island for about 18 hours and left almost the entire population without power.
Irizarry and Acosta also included an estimate of deaths caused by Hurricane Katrina in their study, which struck mainly Louisiana in 2005.
The statistical model used by Irizarry and Acosta coincides with the estimate of deaths associated to Katrina. They place it at 1,832, one less than the number mainly used as the official figure.
For the researchers, after Hurricane Georges - in contrast to the response to Katrina in Louisiana -, in Puerto Rico "there was no systematic effort to improve the power grid or its fragile healthcare system."
💬See 0 comments